The deubiquitinating enzyme USP17 is essential for GTPase subcellular localization and cell motility

Abstract

Deubiquitinating enzymes are now emerging as potential therapeutic targets that control many cellular processes, but few have been demonstrated to control cell motility. Here, we show that ubiquitin-specific protease 17 (USP17) is rapidly and transiently induced in response to chemokines SDF-1/CXCL12 and IL-8/CXCL8 in both primary cells and cell lines, and that its depletion completely blocks chemokine-induced cell migration and cytoskeletal rearrangements. Using live cell imaging, we demonstrate that USP17 is required for both elongated and amoeboid motility, in addition to chemotaxis. USP17 has previously been reported to disrupt Ras localization and we now find that USP17 depletion blocks chemokine-induced subcellular relocalization of GTPases Cdc42, Rac and RhoA, which are GTPases essential for cell motility. Collectively, these results demonstrate that USP17 has a critical role in cell migration and may be a useful drug target for both inflammatory and metastatic disease.

Figure 1. USP17 expression is required for chemotaxis.

Human PBMCs (a,b) or HeLa cells (c,d) were stimulated with CXCL12 (100 ng ml⁻¹) for the indicated times. USP17 mRNA expression was assessed by RT–PCR (a,c) and USP17 protein was determined by immunoprecipitation (IP), followed by immunoblotting (IB) (b,d). In mRNA images, β-actin was used as a loading control, whereas γ-tubulin was the control for western blotting. (e) Jurkat T cells were serum starved and then stimulated with CXCL12 (100 ng ml⁻¹) for 0, 5, 10 and 15 min. USP17 mRNA was analysed by RT–PCR. glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA was also analysed as a loading control. Human neutrophils extracted from blood (f) or the monocytic cell line U937 (g) were stimulated with CXCL8 (100 ng ml⁻¹) for the indicated times. USP17 protein was immunoprecipitated then blotted for USP17. Whole-cell lysates were analysed for γ-tubulin as a loading control. (h,i) HeLa cells were transfected with either scrambled shRNA, USP17 shRNA1 or CYLD shRNA (h) or scrambled and two additional shRNAs targeting USP17 (shRNA2 and shRNA3) (i). Cells were left unstimulated (grey bars) or stimulated with CXCL12 (20 ng ml⁻¹) (black bars) and chemotaxis was measured in modified Boyden chambers. RT–PCR was used to determine USP17 mRNA expression and β-actin was used as a loading control. Experiments were conducted in triplicate; values are mean±s.e.m.; and statistical analyses are relative to untreated control by one-way analysis of variance. **P>0.05.

Figure 2. USP17 depletion inhibits chemokinesis.

Hut78 T cells were cotransfected with either scrambled shRNA or USP17 shRNA1 together with pMAX-GFP to determine transfected cells. Cells were allowed to adhere to culture plates coated with a CD11a/leucocyte function associated-1 antibody for 15 min before image acquisition using the IN Cell Imager. Brightfield images were captured every 60 s for 90 min; fluorescent images at 488 nM were captured at 0 and 90 min to confirm transfection. Scale bar, 20 μm. (a) Representative images of cells at 0, 30, 60 and 90 min (scale bar, 20 μm). (b) Cell trajectories where each line indicates an individual cell's migration. (c) A total of 60 cells were scored for distance and velocity and shown graphically. Each point represents one track, and the bar displays the median value. ***P>0.005 as determined by one-way analysis of variance.

Figure 3. USP17 shRNA alters cell morphology.

HeLa cells transfected with scrambled shRNA were left unstimulated (a) or stimulated with CXCL12 (100 ng ml⁻¹) (b). USP17 shRNA1 cells were unstimulated (c) or stimulated with CXCL12 (100 ng ml⁻¹) (d). Brightfield images were captured every 15 s for 10 min by time-lapse microscopy using Leica SP5. Representative images from triplicate experiments are shown. Scale bar, 20 μm. (e) RT–PCR was used to confirm USP17 depletion using mRNA from HeLa cells by examining USP17 and β-actin mRNA expression.

Figure 4. USP17 depletion inhibits cytoskeleton rearrangements.

HeLa cells were transfected with scrambled shRNA or USP17 shRNA1, and stimulated with CXCL12 (100 ng ml⁻¹) for 0, 2 or 5 min. (a) Cells were costained for actin-phalloidin (red), α-tubulin (green) and 4,6-diamidino-2-phenylindole (blue) and representative images captured by the Leica SP5 microscope. Arrows depict membrane ruffles; scale bar, 20 μm. (b) Quantification of membrane ruffles of triplicate experiments is represented graphically. Black bars are scrambled shRNA-, dark grey bars are USP17 shRNA1- and light grey are CYLD shRNA-transfected cells. Values are mean±s.e.m. ***P>0.005, *P>0.01; NS, not significant as determined by analysis of variance.

Figure 5. Knockdown of USP17 inhibits GTPase localization and activation.

HeLa cells were transfected with scrambled shRNA or USP17 shRNA1 along with GFP-tagged Cdc42 (a), Rac1 (b) and RhoA (c). Cells were either left unstimulated or stimulated with CXCL12 (100 ng ml⁻¹) for 5 min and captured by confocal microscopy (Leica SP5). Arrows indicate sites of plasma membrane localization. Scale bar, 20 μm. (d) Cells were transfected as in a–c, stimulated with CXCL12 (100 ng ml⁻¹) for 0, 0.5, 2 or 5 min and analysed for GTPase activation by performing pulldown assays for the indicated GTPase. Pulldowns and whole-cell lysates were immunoblotted with Cdc42, Rac1 or RhoA antibodies as indicated. Data shown are representative of three experiments.

Figure 6. GTPase mislocalization at the leading edge by USP17 depletion.

HeLa cells were transfected with GFP-tagged Cdc42 (a) or RhoA (b) and either scrambled shRNA or USP17 shRNA1 as indicated. Cells were serum starved and either left unstimulated or stimulated with CXCL12 for 5 min. Cells were then fixed and stained for phalloidin-actin (red) and visualized by confocal microscopy. Scale bar, 10 μm. Arrows indicate areas of actin and GTPase plasma membrane colocalization. (c) HeLa cells were transfected with GFP-tagged RhoA along with either scrambled shRNA or USP17 shRNA1 and cells were serum starved for 12 h before imaging. At least ten brightfield and green TIRF images were captured per condition before and post stimulation with CXCL12 (100 ng ml⁻¹), in triplicate experiments. Fluorescence was scored on the basis of green TIRF per cell; values are mean±s.e.m.; and statistical analyses are relative to scrambled shRNA control by one-way analysis of variance. ***P>0.005. Grey bars indicate unstimulated cells, whereas black bars are CXCL12-stimulated cells.

Figure 7. USP17 depletion results in blunted downstream signalling.

HeLa cells were transfected with either scrambled shRNA or USP17 shRNA1 and stimulated with CXCL12 (100 ng ml⁻¹) for 0, 0.5, 2 or 5 min. Protein was collected and whole-cell lysates were immunoblotted for pPAK, total PAK, active β1 integrin, total β1 integrin, pERK1/2 and total ERK1/2 as indicated.